Phase transition phenomena in an iron phosphate glass and thermal stability criteria
A.Djeridi, A. Benmounah
Research Unit for Materials, Processes and Environment University M' Hamed Bougara of Boumerdes
Boumerdes, Algeria a_djeridi@univ-boumerdes.dz
N. Kamel, D. Moudir, S.Kamariz, Y.Mouheb Nuclear Techniques Division
Algiers Nuclear Research Centre Algiers, Algeria
Abstract— The thermal stability of an iron phosphate glass dedicated to radioactive waste confinement is a crucial characteristic for the integrity of the waste package. In this study, we investigate the thermal stability criteria defined by Hruby, Weinberg and Lu/Liu by measuring the glass transition temperature (Tg), the crystallization temperature (Tc), and the melting temperature (Tm) of an iron phosphate glass loaded with radioactive waste. These temperatures decrease when loading the glass with the waste: Tg decreases from 623.48 °C to 580.04 °C, Tcfrom 766.34 to 679.26 °C and Tmfrom 932.76 °C to 919.75 °C.
Thermal stability criteria show satisfactory values, in the interval 0.14 – 0.69.
Keywords—Iron phosphate glass; Thermal stability;
Radioactive waste; XRD; DTA;SEM I. INTRODUCTION
Phosphate glasses are interesting for their applications in both scientific and technological fields due to their specific physical properties such as both low melting and glass transition temperatures[1-9], high thermal expansions coefficients[4-9], high ultra-violet transmission, high electrical conductivity and compositional flexibility[7−9], Therefore, they are useful for a wide range of applications such as laser hosts[10], biomedical devices[11], solid electrolytes[12] and materials for nuclear waste immobilization[13]. However, their relatively poor chemical durability makes them unsuitable for practical applications[14-16], especially in the field of nuclear waste immobilization. It was reported [16-22]
that the introduction of oxides such as SnO, PbO, ZnO, Cr2O3
and Fe2O3, results in the formation of Sn-O-P, Pb-O-P, Zn-O- P, P-O-Cr and P-O-Fe bonds, and leads to the improvement in the chemical durability of the modified phosphate glasses.
The physical and chemical properties of phosphate glasses can be optimized by controlling the melting conditions and chemical composition. The poor chemical durability of phosphate glasses can be significantly enhanced by the addition of Fe2O3[23−25]. As a result, the iron phosphate glasses (IPG) are of great interest for several technological and biological applications,[26,27]. Iron phosphate glass is considered as an alternate glass for the immobilization of High Level Wastes (HLW) due to its salient features like good glass forming ability, very high waste loading, a favorable chemical stability[28,29], and a low melting temperature typical
between 950 and 1100˚C [30]. Among the various compositions of IPG, the glass with Fe/P ratio of 0.67 is found to be chemically the most durable[31].
The structure of the phosphate glass is based on corner- sharing PO4 tetrahedra which form chains, rings or isolated PO4groups. With the addition of Fe2O3to phosphate glass, the P−O−P bonds are replaced by more chemically durable P−O−Fe2+ and/or P−O−Fe3+ bonds[24,32]. In these glasses, iron commonly exists in two valence states: Fe2+and Fe3+, the Fe2+ ⇔ Fe3+equilibrium depending upon the melting atmosphere and the batch composition[33,34].
In this study, we investigate an iron phosphate glass system with the composition of: 25 % Fe2O3- 57 % P2O5- 8 % B2O3- 8 % Na2O - 2 % As2O3. The glass is synthesized by a double melting method at 1100°C, and is loaded by a Ce-rich complex nuclear waste mixture, containing over 25 elements, which was determined according to that of a simulated fast reactor waste[35], and a spent electrolyte used for pyrochemical reprocessing[36,37]. Both the radioactive waste loaded and non-loaded glasses are characterized by X-Ray Diffraction (XRD), Scanning Electronic Microscopy (SEM), Differential Thermal Analysis (DTA), glass density and molar volume. The glasses stability is assessed by various parameters proposed by Hruby (KH=(Tc- Tg)/(Tm- Tc) [38], Weinberg (KW=(Tc- Tg)/Tm) [39] and Lu/Liu (KLL= Tc/(Tg+ Tm)) [40].
II. EXPERIMENTAL
The glasses are prepared using the method described by Mesko et al. [41] which consisted in two melting/casting steps.
The oxides reagents are dried, then milled in a Retsch Gmbh 5657 automatic agath mortar, till the particles grains sizes are lower than 20 µm.
The IPG glass is prepared by weighing and mixing thoroughly the deferent reagents listed in table 1, such that Fe/P atomic ratio is fixed to 0.67. P2O5is combined with the other components due to its hygroscopic nature. The mixture is homogenized in an Automatic Sieve Shaker D403 during 5 h, to achieve a good particles’ dispersion.
Table 1: Composition of and IPG and IP20RW Oxydes IPG (wt.%) RW (wt.%) IPG20RW (wt.%)
Ag2O 0.256 0.051
Al2O3 2.562 0.512
As2O3 2.000 0.000 1,600
B2O3 8.000 0.000 6.400
BaO 1.367 0.273
CaF2 0.854 0.171
CaO 3.246 0.649
CdO 0.085 0.017
CeO2 49.539 9.908
CoO 0.342 0.068
Cr2O3 0.512 0.102
Cs2O 2.904 0.581
Fe2O3 25.000 11.958 22.392
Gd2O3 0.342 0.068
K2O 0.512 0.102
La2O3 0.683 0.137
MgO 6.833 1.367
MnO2 0.051 0,010
MoO3 5.125 1.025
Na2O 8.000 0.000 6.400
Nd2O3 3.075 0.615
NiO 0.854 0.171
P2O5 57.000 0.000 45.600
Pr6O11 0.854 0.171
Rb2O 0.171 0.034
SnO2 0.137 0.027
SrO 0.342 0.068
TiO2 0.068 0.014
U3O8 0.495 0.099
Y2O3 3.758 0.752
ZnO 0.854 0.171
ZrO2 2.221 0.444
Total (wt.%) 100.000 100.000 100,000
The powders mixture is cooled at 1100°C during 2h, in a high purity alumina crucible in a HF1800 Carbolite furnace, with a heating step of 10°C/min. The melted glasses are poured in water, grinded with a heavy alloy mortar; throw a cotton fiber tissue, to avoid glasses contamination, then milled in a Retsch Gmbh 5657 automatic agath mortar. The grinded glasses are re-melted with the same heating cycle at 1100 °C for 1 h.
The IPG glass loaded by a Ce-rich complex nuclear waste mixture (IPG20RW) was prepared similarly by mixing 80% of IPG with 20% of waste components listed in table 1, with Fe/P ratio maintained equal to 0.67. A batch of 100 g is melted twice at 1050°C during 2h, with a heating step of 10°C/min.
Between the two melting steps, the melts are poured in water, then milled in the Retsch Gmbh 5657 mortar. They are casted into steel cylindrical molds and annealed at 500°C during 1h 30 min to avoid glass internal stress. The formed glasses are cut into fine slices for properties measurements.
The samples densities are measured on grinded powders samples at room temperature by Archimedes method using water as immersion liquid. The estimated error on is about
±0.01 g cm-3.
The volume filled by one mole of glass, Vm, is calculated using the mathematical relation:
V =∑ (1)
where ci is the i oxide molar concentration; Mi the i oxide molar weight, and d the experimental density.
The glass oxygen molar volume, is the volume filled by one mole of oxygen. It is calculated using the following formula:
=
∑∑ (2)
where ni is the i oxide oxygen stoichiometry[42].
The glass transition temperature (Tg), the crystallization temperature (TC) and the melting temperature (Tm) are calculated by differential thermal analysis (DTA), with a heating rate of 5 °C/min-1, using a Netzsch sta 409PC/PG thermal analyzer. The measurements are performed in the temperature range of 20 to 1200°C.
X-ray diffraction (XRD) analysis is carried out with a Philips X’Pert PRO apparatus, from 2 to 90° at a scanning rate of 0.02° 2θ/min. Phases identification is performed using a Philips X’Pert plus 2004 software, using COD database[43].
The glasses microstructure is revealed by SEM (SEM) analysis on transversal sections of the samples, using a Philips XL30 microscope.
III. RESULTS AND DISCUSSONS A. Density and molar volume measurements
The densities and molar volumes of the both IPG and IPG20%RW glasses are presented in Table 2.
The density of IPG20%RW waste form is 3.178 g/cm3, whereas that of IPG 2.888 g/cm3, and the molar volume of IPG20%RW waste form is 39.122 cm3/mol, whereas that of IPG 42.635 cm3/mol. The effect of waste loading in the matrix is traduced by a rise in both the glass density and molar volume.
TABLE 2 : ARCHIMEDES DENSITIES AND MOLAR VOLUMES OF THE SYNTHESIZED GLASSES.
Glasses d (g/cm3) Vm(cm3·mol-1) (cm3·mol-1)
IPG 2.8877 42.635 11.614
IPG20SRW 3.1779 39.122 11.790
Ref BGS [44] 2,75 B. Thermal stability
Tg, Tc and Tm glasses characteristics temperatures are determined for IP20RW and compared with that of IPG to understand the effect of RW loading during the formation of the glass. DTA diagrams are depicted in Fig. 1. The results of the allotropic transformations temperatures are given in Table 3.
Fig.1. DTA diagrams of the studied glasses
TABLE 3: COMPARISON OF Tg, Tc AND Tm VALUES OF THE PREPARED GLASSES WITH THE VALUES OF THE LITTERATURE
Glasses Tg Tc Tm KH
±10-3 KW
±10-3 KLL
±10-3
IPG 623.48 766.34 932.75 0.86 0.15 0.49
IPG20SRW 580.04 679.26 929.43 0.41 0.11 0.45 IPG in K.
Josef et al[35] 509.85 694.85 924.85 0.80 0.15 0.49 IPG20%RW
in Josef et al
[35] 511.85 676.85 917.85 0.69 0.15 0.48
S. Li et al [45] 528 to
579 747 to
876 --- --- --- ---
Q. Liao et al
[46] 526 to
529 625 to
627 --- --- --- ---
K. Kofuji et
al [37] 488 to
522 --- --- --- --- ---
Tg value of the pure glass is of 623.48°C and decreases to 580.04°C when loading the glass with the complex oxides mixture of RW. However, this value is acceptable since it is over the waste package temperature during storage (550°C).
Both the crystallization and melting temperatures decreases also when loading the glass with the RW powder mixture: one can remark a Tc decrease of 87°C from 766.34 to 679.26°C,
from the pure to the RW loaded glass; and a little decrease of 3°C of Tm from 932.75 to 929.43°C.
S. Li et al. [45] studied the influence of the content of MgO, CaO, SrO on the glass stability for a glass with the composition of : (60P2O5–40Fe2O3)-x RO (RO=MgO, CaO, SrO, x=5,10,15,20 mol%) and found that the both Tg and Tc glass temperatures grow up by increasing MgO and CaO glass contents (Tg varies from 528 to 579°C, and Tc from 747 to 876 °C) contrary to the glasses where SrO content rises. These glasses slump dramatically, and their thermal stability deteriorates as well in this system.
Q. Liao et al. [46] synthesize a RW loaded IPG glass with Fe contents varying from 18.6 to 24.6 wt.% Fe2O. Tg increases slightly from 526 to 529°C, and Tc from 625 to 627°C. These values are lower than the values found for the present studied glass. They are close to those given by S. Li et al. [10]. This is due to the less content in alkali elements in the present studied glass.
K. Joseph et al. [35] synthesize a IPG glass containing 20 % of simulated fast reactor waste. They found that Tg slightly increases from 509.85 to 511.85 °C and Tc decreases slightly from 694 to 676°C, and Tm from 924 to 917°C. Except for Tc which is very close to Tc for the present RW loaded glass, these values are also lower to our values. The similarity in Tc values is attributed to the similarity of chemical compositions of the studied glasses.
H. Kofuji et al. [37] studied an IPG glass as an alternative waste form for high level waste generated from pyrochemical reprocessing. Optimization experiments of glass composition were carried out to investigate the effect of additional transition metal oxides in the glass. The glas transition temperature varies from 488-522°C with the chemical composition.
In order to evaluate the glass stability, it is essential to assess the glass resistance to crystallization during re-heating. It is expressed by GS criterium [47]. GS can be estimated from Tg, Tcand Tmvalues (Table 2).
Hruby (KH=(Tc- Tg)/(Tm- Tc), Weinberg (KW=(Tc- Tg)/Tm) and Lu/Liu (KLL= Tc/(Tg+ Tm)) criteria are used to assess GS (Table 2).
These criteria are similar to that of K. Joseph et al. [9] proving the thermal stability of the present studied IPG glass.
The larger are KH, KW and KLL values, the greater would be the glass stability against crystallization. The values obtained for silicate glasses [48,49] in the literature are 0.14–0.69.
However, these authors [48,49] used the melting temperature (Tm) for calculating GS.
A similarity in the estimated GS parameter between IPG and IP20RW glasses suggests that the glass stability is not altered by the addition of 20 wt.% of simulated wastes.
C. XRD and SEM Analyses
The XRD spectra confirm the amorphous structure of the glasses, with no residual crystalline phase (fig. 2).
0 200 400 600 800 1000 1200
0,0 0,3 0,6 0,9 1,2 1,5 1,8
Heat flow (µV/mg)
T (°C) Iron phosphate glass Iron phosphate glass with RW
Fig.2. XRD diagrams of the studied glasses.
Typical SEM micrographs of both pure and RW loaded glasses are depicted on Fig.3.
The glasses microstructure is revealed by both the bulk and surface SEM observations. The micrographs confirm the homogeneity of the glasses for the different Fe contents. The SEM analysis confirms the XRD results, and thus the amorphous nature of the products. No crystalline micro-grains can be visualized with these spectroscopic methods.
3.a.
Fig..3. A typical SEM micrograph of the studied glasses (a) the IPG and (b)3.b of IPG20RD.
IV.CONCLUSION
The thermal stability of an IPG glass dedicated to radioactive waste confinement is assessed by a DTA analysis. A glass system with the composition of: 25 % Fe2O3- 57 % P2O5- 8 % B2O3 - 8 % Na2O - 2 % As2O3 is synthesized by a double melting method at 1100°C, and is loaded by a Ce-rich complex nuclear waste mixture, containing over 25 elements.
Both the IPG and IPG20%RW are characterized by XRD, SEM, and DTA analysis. The effect of waste loading in the matrix is traduced by a rise in the glass density and molar volume from 2.888 to 3.178 g/cm3, for density and from 39.122 to 42.635 cm3/mol for the molar volume.
Tg value of the IPG glass is of 623.48°C and decreases to 580.04°C when loading the glass with the radioactive waste mixture. However, this value is acceptable. Both Tc and Tm temperatures decreases also when loading the glass with the RW powder mixture. The glasses stability is assessed by Hruby, Weinberg and Lu/Liu criteria. These thermal stability criteria show satisfactory values, in the interval 0.14 – 0.69.
References
[1] L. M. Sanford and P. A. Tick, US Patent 4.314.031, 1982.
[2] P. Y. Shih, S. W. Yung and T. S. Chin, “Thermal and Corrosion Behavior of P2O5-Na2O-CuO Glasses,” Journal of Non-Crystalline Solids, Vol. 224, No. 2, 1998, pp. 143- 152. doi:10.1016/S0022- 3093(97)00460-2
[3] P. Y. Shih, S. W. Yung, C. Y. Chen, H. S. Liu and T. S. Chin, “The Effect of SnO and PbCl2on Properties of Stannous Chlorophosphate Glasses,” Materials Chemistry and Physics, Vol. 50, No. 1, 1997, pp. 63- 69. doi:10.1016/S0254-0584(97)80185-X
[4] T. Y. Wei, Y. Hu and L. G. Hwa, “Structure and Elastic Properties of Low-Temperature Sealing Phosphate Glasses,” Journal of Non- Crystalline Solids, Vol. 288, No. 1-3, 2001, pp. 140-147.
doi:10.1016/S0022-3093(01)00612-3
[5] H. Niida, M. Takahashi, T. Uchino and T. Yoko, “Preparation and Structure of Organic-Inorganic Hybrid Precursors for New Type Low- Melting Glasses,” Journal of Non-Crystalline Solids, Vol. 306, No. 3, 2002, pp. 292-299. doi:10.1016/S0022-3093(02)01188-2
[6] H. S. Liu, P. Y. Shih and T. S. Chin, “Thermal, Chemical and Structural Characteristics of Erbium-Doped Sodium Phosphate Glasses,” Physics and Chemistry of Glasses, Vol. 37, 1996, p. 227.
[7] R. K. Brow, J. Non-Cryst. Solids 2000, 263−264, 1.
[8] B. C. Sales, L. A. Boatner, Science 1984, 226, 45.
[9] Y. Abe, H. Hosono, Inorganic Phosphate Materials, Edited by T.
Kanazawa. Elsevier, Amsterdam, 1989, p. 124.
[10] T.T. Fernandez, P. Haro-Gonzales, B. Sotello,M. Hernandez, D. Jaque, P. Fernandez, C.Domengo, J. Siegel, J. Solis, Opt. Lett. 38 (2013) 5248–
5251.
[11] K. Takada, Acta Mater. 61 (2013) 759–770.
[12] J.K. Christie, R.I. Ainsworth, D. Tomaso, N.H. de Leeuw, J. Phys.
Chem. B 117 (2013) 10652–10657.
[13] P. Sengupta, J. Hazard. Mater. 235 (2012) 17–28.
[14] M. R. Reidmeyer, M. Rajaram and D. E. Day, “Preparation of Phosphorus Oxynitride Glasses,” Journal of Non- Crystalline Solids, Vol. 85, No. 1-2, 1986, pp. 186-203. doi:10.1016/0022-3093(86)90090- 6
[15] H. Yung, P. Y. Shih, H. S. Liu and T. S. Chin, “Nitridation Effect on Properties of Stannous-Lead Phosphate Glasses,” Journal of the American Ceramic Society, Vol. 80, No. 9, 1997, pp. 2213-2220.
doi:10.1111/j.1151-2916.1997.tb03110.
[16] J. L. Rygel and C. G. Pantano, “Synthesis and Properties of Cerium Aluminosilicophosphate Glasses,” Journal of Non-Crystalline Solids,
20 40 60 80 100
0 20 40 60 80 100 120 140 160 180 200 220 240
Intensity (a.u)
2 (°)
IPGIPG20% RW
Vol. 355, No. 52-54, 2009, pp. 2622-2629.
doi:10.1016/j.jnoncrysol.2009.09.004
[17] P. A. Bingham and R. J. Hand, “Sulphate Incorporation and Glass Formation in Phosphate Systems for Nuclear and Toxic Waste Immobilization,” Materials Research Bulletin, Vol. 43, No. 7, 2008, pp.
1679-1693. doi:10.1016/j.materresbull.2007.07.024
[18] S. Ray, X. Fang, M. Karabulut, G. K. Marasinghe and D. E. Day,
“Effect of Melting Temperature and Time on Iron Valence and Crystallization of Iron Phosphate Glasses,” Journal of Non-Crystalline Solids, Vol. 249, No. 1, pp. 1-16. doi:10.1016/S0022-3093(99)00304-X [19] P. Y. Shih and T. S. Chin, “Preparation of Lead-Free Phosphate Glasses
with Low Tg and Excellent Chemical Durability,” Journal of Materials Science Letters, Vol. 20,No. 19, 2001, pp. 1811-1813.
doi:10.1023/A:1012551603964
[20] C. M. Shaw and J. E. Shelby, “Effect of Lead Compounds on the Properties of Stannous Fluorophosphate Glasses,” Journal of the American Ceramic Society, Vol. 71, No. 5, 1988, pp. C252-C253.
doi:10.1111/j.1151-2916.1988.tb05071.x
[21] I. W. Donald, “Preparation, Properties and Chemistry of Glass- and Glass-Ceramic-To-Metal Seals and Coatings,” Journal of Materials Science, Vol. 28, No. 11, 1993, pp. 2841-2886.
doi:10.1007/BF00354689
[22] A. Šantić and A. Moguš-Milanković, “Charge Carrier Dynamics in Materials with Disordered Structures: A Case Study of Iron Phosphate Glasses,” Croatica Chemica Acta, Vol. 85, No. 3, 2012.
doi:10.5562/cca1989
[23] D. E. Day, Z. Wu, C. S. Ray, P. Hrma, J. Non-Cryst. Solids 1998, 241, 1.
[24] X. Yu, D. E. Day, G. J. Long, R. K. Brow, J. Non-Cryst. Solids 1997, 215, 21.
[25] X. Fang, C. S. Ray, A. Moguš-Milanković, D. E. Day, J. Non-Cryst.
Solids 2001, 283, 162.
[26] B. C. Sales, L. A. Boatner, Radioactive Waste Forms for the Future, Eds. W. Lutze, R.C. Ewing, North-Holland, Amsterdam, 1988, p. 193.
[27] T. Jermoumi, M. Hafid, N. Niegisch, M. Mennig, A. Sabir, N. Toreis, Mater. Res. Bull. 2002, 37, 49.
[28] D. E. Day, Z. Wu, C. S. Ray, P. Hrma, J. Non-Cryst. Solids, 241, 1 (1998).
[29] Kitheri Joseph, K. V. Govindan Kutty, P.Chandramohan, P. R.
Vasudeva Rao, J.Nucl. Mater., 384, 262 (2009).
[30] S. Aqdim, H. El Sayouty, B. Elouad and J. M. Greneche, “Chemical Durability and Structural Approach of the lass Series (40-y) Na2O- yFe2O3-5Al2O3-55P2O5-by IR, X-Ray Diffraction and Mössbauer Spectroscopy,” IOP Conference Series: Materials Science and Engineering, Vol. 28, 2012, Article ID:012003.
[31] M. Karabulut, G.K. Marasinghe, C.S. Ray, D.E. Day, O. Ozturk, G.D.
Waddill, J. Non-Cryst . Solids 249, 106 (1999).
[32] R. K. Brow, C. M. Arens, X. Yu, D. E. Day, Phys. Chem. Glasses 1994, 35, 132.
[33] C. S. Ray, X. Fang, M. Karabulut, G. K. Matasinghe, D. E. Day A., J.
Non-Cryst. Solids 1999, 249, 1.
[34] G. K. Marasinghe, M. Karabulut, C. S. Ray, D. E. Day, M. G.
Shumsky,W. B. Yelon, C. H. Booth, P. G. Allen,
[35] Kitheri Joseph,, R. Asuvathraman, R. Venkata Krishnan , T.R.
Ravindran , R. Govindaraj,K.V. Govindan Kutty , P.R. Vasudeva Rao, Iron phosphate glass containing simulated fast reactor waste:
Characterization and comparison with pristine iron phosphate glass Journal of Nuclear Materials 452 (2014) 273–280
[36] Japan Atomic Energy Agency, Feasibility Study on Commercialized Fast Reactor Cycle Systems Technical Study Report of Phase II - (2) Nuclear Fuel Cycle Systems-, JAEA Research 2006-043 (2006) pp.368 (in Japanese).
[37] H. Kofuji, T. Yanob, M. Myochina, K. Matsuyama, T. Okita, S.
Miyamotoc, Chemical durability of iron-phosphate glass as the high level waste from pyrochemical reprocessing, Procedia Chemistry 7 ( 2012 ) 764 – 771
[38] A. Hruby, Czech. J. Phys. B22 (1972) 1187–1192.
[39] M.C. Weinberg, J. Non-Cryst. Solids 167 (1994) 81–88.
[40] Z.P. Lu, C.T. Liu, Intermetallics 12 (2004) 1035–1043.
[41] M.G. Meskoa, D.E. Daya, B.C. Bunkerb, Immobilization of CsCl and SrF2 in iron phosphate glass, Waste Management 20 (2000) 271-278 [42] A. Quintas, O. Majérus, D. Caurant, J.L. Dussossoy, P. Vermaut,
Crystallization of a rare earth-rich aluminoborosilicate glass with varying CaO/Na2O ratio, Journal of the American Ceramic Society 90 (3) (2007) 712-719
[43] JCPDS data Philips X’Pert High Score Package, Diffraction data CD- ROM, International Center for Diffraction Data, Newtown Square PA, 2004.
[44] Inagaki, et al., Status of Studies on HLW Glass Performance for Confirming Its Validity in Assessment, Journal of Nuclear Fuel Cycle and Environment, Vol.10 No.1-2 (2004) 69 (in Japanese).
[45] S. Li, H. Liu, F. Wu, Z. Chang, Y. Yue, Effects of alkaline-earth metal oxides on structure and properties of iron phosphate glasses, J. Non- Cryst. Solids, Vol. 434, pp. 108-114, 2016.
[46] Qilong Liao, Fu Wang, Kuiru Chen , Sheqi Pan , Hanzhen Zhu , Mingwei Lu , Jianfa Qin, FTIR spectra and properties of iron borophosphate glasses containing simulated nuclear wastes , Journal of Molecular Structure 1092 (2015) 187–191]
[47] J.E. Shelby, Introduction to Glass Science and Technology, second ed., The Royal Society of Chemistry, RSC, 2005.
[48] A.A. Cabral, A.A.D. Cardoso, E.D. Zanotto, Glass-forming ability versus stability of silicate glasses, J. Non-Cryst. Solids 320 (2003) 1-8.
[49] S.E. Lin, Y.R. Cheng, W.C.J. Wei, J. Eur. Ceram. Soc. 31 (2011) 1975.
